ERA5

Homepage
Download location: Copernicus

Download procedure (info; YouTube video series):

• Request fields, region, and timespan at download location

• Download locally, or use CDSAPI, python-based approach.

  • YouTube tutorial for CDSAPI

  • Example of CDSAPI python script with for loop

  • Use online selection tool at download location to check variable names, as well as header info and product type for specific dataset.

• Option to download GRIB file and convert to NetCDF (how to; simplest approach is using cdo) OR download directly as netcdf (preferred)

• A downloadable grid file is not available for ERA5. See here for information about the grid. Code to evaluate area for a regular lat/lon grid is available in my utils package here (in the geo module).

Notes¶

Differing timestamps across variables¶

Downloading ERA5 monthly data, heat fluxes from the ocean surface have a timestamp at 0600 whereas state variables like air and surface have a timestamp at 0000. A preprocess step using the floor category of the datetime accessor in xarray allows them to be loaded together:

import xarray as xr
basedir = '/gws/nopw/j04/co2clim/datasets/ERA5/monthly/'
filename = 'era5.1x1.monthly.*.*.nc'
def preprocess(ds):
    time = ds.valid_time.dt.floor("D")
    return ds.assign_coords(valid_time=time)
ds = xr.open_mfdataset(basedir+filename,preprocess=preprocess)